Let’s just call this what it is.
I hit a reset at 53% usage, which means 47% of what I paid for just vanished. Not because I used it. Not because I mismanaged it. Just because the clock decided to cut the cycle short.
My UI literally says it: 11 hours 21 minutes stolen by early reset.
So this isn’t a “feels like” situation, its the system itself, through true usage measureents admitting it.

I budget my usage like a normal person (and this week, I left a lot for Monday, which was to be my 'big job day'):
Usually, I spread it across the full cycle, lean on caching to stretch it, leave headroom for spikes. That only works if the time window is real.

But it’s not.

Instead, what I actually get is: “Here’s your full quota… but maybe only for half the time. Good luck.”
So in this case:
Paid for: 100%
Actually usable: 53%
Straight-up lost: 47%

And here’s the worst part:
That effectively means I’m paying almost double for what I actually got to use.
Like, do the math: If I only get to use ~53% of what I paid for, then my real cost per usable unit is basically 1.9x what it should be. That’s not a rounding error. That’s a completely broken value model.

Why this screws my workflow specifically
My whole setup assumes:
I have X capacity over Y time that I can choose to pace usage intelligently or choose when to spend heavy. But with early resets, it turns into: “You had capacity… you just didn’t use it fast enough before we took it back.” Cool. Love that... not. There’s no way to “make that up” either. I can’t go back in time and burn it faster.

The real issue
This isn’t about usage. It’s about unreliable delivery. If quota is tied to time, then the time window has to be consistent. Otherwise: the quota isn’t actually guaranteed, the budgeting model is meaningless, & planning ahead is basically pointless

Here's My Truth
I didn’t underuse my quota.
My quota was taken away from me early.

Those are not the same thing.

And losing 47% of paid capacity because of timing alone is… yeah, not exactly a great system.